Discrimination in Ingestion of Protistan Prey by Larval Crabs

MARINE ECOLOGY PROGRESS SERIES Vol. 222: 155–162, 2001 Published November 5 Mar Ecol Prog Ser

Discrimination in ingestion of protistan prey by larval crabs

Shawn Hinz, Stephen Sulkin*, Suzanne Strom, Jill Testermann

Shannon Point Marine Center, 1900 Shannon Point Road, Anacortes, Washington 98221, USA

ABSTRACT: We determined the incidence of ingestion of 4 autotrophic dinoflagellates and 1 heterotrophic dinoflagellate by first stage larvae of 4 species of crabs. Crab species were 2 winter spawning brachyurans Cancer magister and C. oregonensis, 1 summer spawning brachyuran Hemigrapsus oregonensis, and 1 anomuran Rhinolithodes wosnessenskii. Autotrophic dinoflagellate prey were Prorocentrum micans, which sustain survival of crab larvae in laboratory culture, and 2 species of Alexandrium spp. that do not. P. micans were ingested by virtually all larvae of all 4 crab species, while both toxic and non-toxic strains of Alexandrium were almost never ingested. Results of rearing experiments generally confirmed that larvae were receiving no nutritional contribution from Alexan- drium spp. prey. When brachyuran larvae were presented with mixtures of P. micans and Alexan- drium spp. in defined ratios, virtually all larvae ingested both types of algal prey. Suspending Alexan- drium cells in P. micans exudate did not enhance their ingestion nor did suspending P. micans in Alexandrium exudate reduce ingestion. Ingestion of plastic beads was low (<12%) except when offered in combination with P. micans cells (58%). H. oregonensis larvae ingested the heterotrophic dinoflagellate Noctiluca scintillans that had previously fed on either P. micans or one of the toxic Alexandrium strains, with no apparent preference. Results suggest the presence of a positive ingestion stimulus provided by P. micans and N. scintillans, but its absence in Alexandrium spp. Absence of ingestion of Alexandrium was not related to the presence of toxins. The ingestion stimulus appears to reside on the prey cell surface. Although crab larvae appear able to discriminate among algal prey, non-discriminate feeding seems likely to occur in mixed prey assemblages in which at least some prey possess the positive ingestion cue, perhaps permitting rapid ingestion of available particles when dense prey patches are encountered in an otherwise sparse prey environment.

KEY WORDS: Protists · Crabs · Larvae · Nutrition · Prey discrimination

Resale or republication not permitted without written consent of the publisher

INTRODUCTION moting good survival during later zoeal stages (e.g. Anger & Dawirs 1981, Staton & Sulkin 1991, Hartman Larvae of most species of brachyuran crabs are & Sulkin 1999). Although brachyuran larvae have planktotrophic; that is, they must feed on particulate been thought to require relatively high densities of sources of nutrition in the plankton to develop success- motile animal prey, Lehto et al. (1998) and Sulkin et fully to metamorphosis. In addition to needing a source al. (1998a,b) have shown that larvae of the crabs Can- of energy to sustain metabolism, most species must cer spp. and Hemigrapsus oregonensis ingest both obtain specific nutrients from the diet to maximize sur- autotrophic and heterotrophic protists and derive vival and development rate (Sulkin & van Heukelem nutritional benefit from them. They speculated that 1980, Levine & Sulkin 1984). while such diets alone may not be sufficient to sustain Obtaining a source of nutrition early during the first larval development, protists may be an important larval (zoeal) stage is particularly important in pro- source of nutrition, sustaining first-feeding larvae until they encounter sufficient densities of suitable *Corresponding author. E-mail: [email protected] metazoan prey. This capacity to utilize the carbon

sources of the microbial food web may be particularly bioassay procedures before and at the conclusion of important since suitable metazoan prey are often the experiments. patchy in distribution and may not be present in suffi- Crab species tested were 3 brachyurans, the 2 winter cient densities to sustain larvae (Damkaer 1977, Paul spawners Cancer magister and C. oregonensis and the et al. 1979, 1989). summer spawner Hemigrapsus oregonensis. An ano- However, while protists such as the autotrophic muran lithodid, Rhinolithodes wosnessenskii, was in- dinoflagellate Prorocentrum micans and the hetero- cluded because its larvae are facultative planktotrophic dinoflagellate Noctiluca scintillans have pro- trophs, able to feed but not requiring an external ved effective at providing nutrition to crab larvae, not source of nutrition to survive. all protists are. For example, the ciliates Euplotes van- Larvae were fed both single and mixed-species algal nus and Parauranema virginianum did not sustain diets as well as combinations of algal cells and plastic development of larval crabs (Sulkin 1975). Further, lar- beads, and algal cells immersed in the freshly pre- vae that were fed the ciliates Strombidinopsis acumi- pared exudates of other phytoplankton species. In natum and an Uronema sp., the dinoflagellates Oxyrr- addition, larvae were fed the heterotrophic dinoflagel- hris marina, and 2 species of Alexandrium did not late Noctiluca scintillans that itself was cultured in the show higher survival than did unfed larvae (Sulkin laboratory on either Prorocentrum micans or Af1719. unpubl. data). It is not clear whether these differences Experimental animals. All 4 crab species were col- among protists as prey for crab larvae are due to selec- lected in ovigerous state in the inland marine waters of tive ingestion on the part of the larvae or to differences the northern Puget Sound basin near the Shannon in biochemical composition of the various prey. Point Marine Center in Anacortes, Washington. Larvae The present study was designed to determine of Cancer magister and C. oregonensis hatch in early whether crab larvae are selective in their ingestion of winter, with larvae of the latter species being generally different protists. We compared the ingestion of an more resistant to nutritional stress (Sulkin & McKeen autotrophic dinoflagellate known to sustain survival of 1999). Larvae of these species are present in the water crab larvae Prorocentrum micans with others that do column during a period of very low primary productiv- not Alexandrium tamarense and A. fundyense. Use of ity and may be particularly dependent on a source of Alexandrium spp. as experimental prey also permitted microbial nutrition during the critical period of early examination of larval ingestion of both toxic and non- development (Sulkin et al. 1998a). Larvae of Hemi- toxic strains of an autotrophic dinoflagellate. grapsus oregonensis hatch in the summer, a period of comparatively high primary productivity and of fre- quent blooms of toxic algae including Alexandrium MATERIALS AND METHODS spp. Rhinolithodes wosnessenskii larvae have not been studied extensively, but typically hatch locally in the Experimental protocol. Experiments were con- early spring (Haynes 1984). Their ingestion of protists ducted to determine whether newly hatched larvae of was of interest because their nutritional needs contrast 4 species of crabs would ingest cells of target protists with those of the 3 brachyuran species. provided individually and in various combinations. Preparation and maintenance of diet treatments. Ingestion was determined by direct observation of lar- Autotrophic dinoflagellates: Prorocentrum micans was val guts as described below. cultured in 1 l polycarbonate bottles in f/2 nutrient me- Diet treatments included 2 autotrophic dinoflagel- dium maintained at 22°C on a 12/12 h light/dark cycle. lates that, although different in shape and volume, are To maintain cells in growth phase and to normalize for approximately the same diameter (35 to 50 µm). Proro- the feeding experiments, cultures were maintained at centrum micans (strain UTEX 1993), previously shown an approximate cell density of 1000 ml–1. Starter cul- to sustain crab larvae in laboratory culture (Lehto et al. tures were not axenic but contained no other algae. All 1998, Sulkin et al. 1998b), was used as a control diet. 3 strains of Alexandrium were cultured using the same Two species of Alexandrium (shown not to sustain lar- procedures. val development in preliminary tests) were used in the Noctiluca scintillans: The heterotrophic dinoflagel- experiments: A. fundyense 1719 (Af1719), a strain shown late Noctiluca scintillans was cultured in 2 l polycar- to be toxic in mouse bioassays, and 2 strains of A. ta- bonate bottles filled with 0.2 µm filtered seawater and a marense, 1 toxic (At118) and 1 non-toxic (At115) (Tee- dilute EDTA-trace metal mix (Gifford 1985). Stock cul- garden & Cembella 1996). All phytoplankton strains tures were fed weekly on Prorocentrum micans. Before were obtained from the National Center for Collection the ingestion experiments, a stock culture was taken off of Marine Phytoplankton (Bigelow Laboratory, West the P. micans diet and starved for 5 d to clear its food Boothbay Harbor, ME). The toxicity of the strains used vacuoles of cells. The starved N. scintillans cells were in the present study was confirmed by standard mouse then divided into 2 groups, 1 fed for 48 h on P. micans Hinz et al.: Larval crab ingestion of protistan prey 157

and the other on Af1719. Both algae were provided to was deemed to have occurred when there were at least N. scintillans at cell densities of 3000 ml–1 from cultures 3 individual algal cells in the gut. In the vast majority of grown to exceed that density. Twenty N. scintillans cases, the guts either contained many cells or were were selected randomly from each treatment for exam- empty. Ingestion tests were identical for all treatments. ination of their food vacuoles to insure that the cultures Larvae used in the experiments had hatched during were actively feeding on the respective algal diets. N. the previous 12 h and had not been fed. For each scintillans cells were concentrated by reverse filtration observation, larvae from at least 2 broods were pooled (80 µm sieve) for feeding to crab larvae. before selection of individual larvae for the experi- Exudate preparations: To determine whether chemi- ment. For each test, 100 larvae were arbitrarily cals exuded by the various algae might have an influ- selected from a large pool of at least several hundred, ence on ingestion behavior, Prorocentrum micans cells were placed in one 80 mm diameter glass bowl, and were suspended in freshly prepared exudates of Af1719 were fed the target diet. At specified observation (toxic) and At115 (non-toxic), while Alexandrium cells of times, 15 larvae were selected haphazardly and ob- both species were suspended separately in exudate from served as described above for evidence of prey cells in P. micans. All 3 cell types were gravity filtered from the their guts. Larvae were discarded after observations. growth medium (10 µm filter) to produce both exudate This experiment was repeated with 3 different groups and isolated cells. P. micans cells were then suspended in of larvae, with treatment comparisons based on the either of the Alexandrium cell exudates and both percentage of larvae feeding in each experiment Alexandrium cells were suspended separately in the P. (n = 3). Ingestion of single diet types and 1:1 ratios was micans exudate. Cell concentrations after suspension in typically observed at 3, 6, 12, 24, 48, and 96 h. Time exudate were approximately 1000 ml–1. Because rapid intervals of observations for other experiments varied lability of exuded Alexandrium toxin was reported by and are specified with the results. Hansen (1989), ingestion observations were made less In the Noctiluca scintillans experiment, a mixture of than 1 min after exudate was isolated. 10 cells fed Prorocentrum micans and 10 fed Af1719 Combination treatments: To determine the effect of was fed to each of 10 Hemigrapsus oregonensis larvae, the presence of 1 algal species on the ingestion of the each placed in a separate 10 mm diameter glass bowl other, defined mixtures were fed to crab larvae. To for 8 h. At the end of 8 h, the remaining N. scintillans produce the mixtures, stock cultures of target species cells were examined and identified as to treatment. were diluted to cell densities of 1000 ml–1 and appro- The experiment was repeated 3 times with different priate volumes of each were combined to produce the groups of larvae. The percentage of each type of N. desired ratio at a total cell density of 1000 ml–1. For scintillans cell ingested by each larva was determined example, a 1:1 ratio of Prorocentrum micans to At118 and compared. was produced by manipulating each stock culture to Larval survival and development rate experiments. 1000 ml–1, then combining equal volumes of the 2 stock To further assess the value of toxic and non-toxic cultures to provide the experimental diet. These 1:1 strains of Alexandrium to crab larval development, lar- ratios of P. micans to Af1719, At115, and At118 were val cultures of the 2 Cancer spp. were maintained from fed to larvae of all 3 brachyuran species. In addition, hatching to molt to zoeal stage 2 on Af1719 and At115. larvae of Hemigrapsus oregonensis were fed P. mi- An unfed treatment and a fed control consisting of cans: Af1719 ratios of 1:2, 1:5, 1:10, 1:15, 1:20, and 1:30. freshly hatched nauplii of the brine shrimp Artemia Additional experiments used plastic beads (Fluo- spp. (Argentemium, Argent Chemical Laboratory, rosprite 25 µm latex beads; Polysciences, Inc., War- Redmont, WA) were also tested simultaneously. Algae rington, PA) either alone or in combination with Proro- were fed to the larvae at a concentration of 1000 cells centrum micans or Af1719 cells at a total cell (particle) ml–1 and Artemia spp. at 8 ml–1. density of 1000 ml–1. Plastic beads were also sus- Newly hatched larvae from at least 2 broods were pended in freshly prepared exudate of either P. micans pooled. Larvae were haphazardly selected from the or Af1719 at a total particle count of 1000 ml–1. Plastic pool and placed in plastic trays containing 12 individ- microsphere experiments were carried out using lar- ual cell wells, 1 larva well–1. Each well contained vae of Hemigrapsus oregonensis only. approximately 4 ml of Alexandrium culture medium Ingestion experiments. Ingestion was determined by containing the algal cells. The unfed larvae were examining the guts of selected larvae for presence of placed into culture medium with no prey present. the target algal cells using an epifluorescence micro- Three trays (each containing 12 larvae) were haphaz- scope. Individual larvae to be examined were removed ardly assigned to each of the 4 treatments and fed by pipet from the feeding culture and placed on a accordingly. Larvae were transferred daily to clean cell microscope slide. Excess water was removed from the wells containing fresh medium with the appropriate slide and the larvae were observed at 40×. Ingestion diet. Mortality and evidence of molting to zoeal stage 158 Mar Ecol Prog Ser 222: 155–162, 2001

2were noted daily. Treatment comparisons were AF1719 at 1000 cells ml–1, 1:1 combination ratio of Pro- based on whether the diet supported development to rocentrum micans:Af1719 at 500 cells ml–1 each (total zoeal stage 2 and, if so, whether there was a difference density of 1000 ml–1), and P. micans only at the reduced in duration of the first zoeal stage. In treatments where dosage of 500 cells ml–1. Larvae were obtained from there was no molting to zoeal stage 2, mean days of 1 female crab and distributed among 3 cell well trays death were calculated and compared. for each diet treatment (n = 36 larvae treatment–1). Lar- Because Rhinolithodes wosnessenskii larvae can de- vae were transferred daily to fresh medium and fed the velop to zoeal stage 2 without feeding, the experimen- appropriate diet. Survival was monitored, with mean tal design for survival and development rate differed days of death compared among the treatments. from that of the 2 Cancer spp. Mortality and percentage of larvae molting on unfed, freshly hatched Arte- mia spp. nauplii and Af1719 (1000 cells ml–1) treat- RESULTS ments were compared on Day 5 of the culture. This day was selected as providing a sufficient feeding time for Ingestion of individual diets dietary effects to be evident, while providing an end- point before 100% molting of unfed larvae. Larvae Prorocentrum micans were ingested by virtually all were obtained from 1 female and distributed among larvae at all observation times in all 4 species tested 9 trays, each containing 12 larvae. Three trays were (Table 1). Cells were present in 90 to 100% of the lar- haphazardly assigned to each diet treatment. Cultures vae at the first (3 h) measurement and guts remained were transferred daily to fresh medium containing the filled with P. micans throughout the 96 h period of the appropriate diet treatment. experiment. Larvae of Hemigrapsus oregonensis were raised By contrast, both non-toxic (At115) and toxic (At118, from hatching on the following diet treatments: unfed, Af1719) strains of Alexandrium were almost never ingested by any species at any observation time Table 1. Mean percentage of larvae that show ingestion of (Table 1). Only in Cancer oregonensis was ingestion of indicated prey types at each observation time. Af1719: Alexandrium cells noted and only of the toxic Af1719 Alexandrium fundyense 1719 (toxic); At115: A. tamarense 115 strain (Table 1). In this case, there was initial ingestion (non-toxic); At118: A. tamarense 118 (toxic); Pm: Prorocen- trum micans. n = 3 for each observation time of cells, with 100% of the larvae feeding by 6 h, fol- lowed by a rapid decline in ingestion such that by 48 h, Crab species Diet treatment no larvae were found to have Af1719 cells in their guts. Time interval Pm At118 At115 Af1719 The effects of raising Cancer oregonensis and C. magister larvae on Prorocentrum micans and Alexan- Cancer oregonensis drium strains generally confirmed the ingestion results. 3900071 6 100 0 0 100 C. oregonensis larvae fed the control diet of freshly 12 100 0 0 75 hatched Artemia spp. nauplii showed 93% survival to 24 100 0 0 34 zoeal stage 2 with a mean stage duration of 10.9 ± 1.1 d 48 95 0 0 0 96 100 0 0 4 (mean ± SE). Unfed larvae and those fed the non-toxic Cancer magister At115 or the toxic Af1719 all died before molting to 395000zoeal stage 2. No significant differences in mean days 6 100 2 0 0 of death were found among these 3 treatments (Table 2; 12 100 0 0 0 24 100 3 3 2 ANOVA, p > 0.05), suggesting that the Alexandrium 48 100 0 0 0 cells were not contributing nutritionally. Although the 96 100 0 0 0 ingestion experiments suggested that larvae did feed Hemigrapsus oregonensis 3 100 0 0 0 Table 2. Cancer spp. Mean ± SE days of death for first stage 6 100 0 0 0 larvae of the 2 species fed the indicated diets. No significant 12 100 0 0 0 differences were found among treatments for either species 24 100 0 0 0 48 100 0 0 0 (ANOVA, p > 0.05). Initial sample size = 36 for each treat- 96 100 0 0 0 ment. See Table 1 for definitions of diet treatments Rhinolithodes wosnessenskii 3 100 – 0 0 Diet Crab species 6 100 – 0 0 C. oregonensis C. magister 12 100 – 0 0 24 100 – 0 0 Af1719 12.8 ± 0.56 5.6 ± 0.54 48 100 – 0 0 At115 13.8 ± 0.73 7.8 ± 0.30 96 100 – 0 0 Unfed 11.6 ± 0.24 5.4 ± 0.19 Hinz et al.: Larval crab ingestion of protistan prey 159

briefly on Af1719 cells, there is no evidence that such Table 3. Rhinolithodes wosnessenskii. Mean percentage mor- feeding resulted in a delay in mortality. tality (A) and percentage molt (B) of stage 1 larvae on Day 5 on indicated diet treatments. Shared letter indicates no signi- Results with Cancer magister larvae were similar. ficant difference (p = 0.05) for each measurement Larvae fed the brine shrimp nauplius control developed successfully to zoeal stage 2 (60% survival; mean A ± SE stage duration 8.4 ± 1.0 d). There was no signifi- Diet Mortality (%) SE Tukey’s HSD cant difference in mean days of death among the unfed and 2 Alexandrium-fed treatments (Table 2; ANOVA, Unfed 26 0.47 a Artemia spp. 10 0.20 b p > 0.05). Again, there is no evidence that sustained Af1719 8 0.61 b exposure to either Alexandrium diet resulted either in B nutritional benefit or toxicity to the larvae. Diet Molt (%) SE Tukey’s HSD Results from the facultative planktotroph Rhinolith- odes wosnessenskii provided an interesting contrast Artemia spp. 65 0.12 a Unfed 32 0.29 b with the 2 Cancer spp. On Day 5, there were signifi- Af1719 17 0.43 c cant differences among treatments in percentage mortality (Table 3; ANOVA on arcsine transformed data, p < 0.05). Results of a Tukey’s honestly significant difference (HSD) test (Table 3) show a higher mortality diet compared with the unfed control. Larvae fed the P. for unfed larvae than for larvae fed either Artemia nau- micans diet at 500 cells ml–1 did not survive to zoeal plii or AT1719. There were also significant differences stage 2 in this experiment. However, adding Af1719 at among treatments in mean percentage molting on 500 cells ml–1 to P. micans at 500 cells ml–1 neither Day 5 (Table 3; ANOVA on arcsine transformed data, accelerated nor delayed mortality. p< 0.05), with feeding on brine shrimp nauplii acceler- ating development, while feeding on Af1719 delaying it compared with the unfed treatment (Tukey’s HSD Exudate experiments test, p = 0.05). These results imply that at least a few of the larvae were feeding on Af1719 either throughout To determine whether the differences in incidence of the experiment in contrast to the direct observations on ingestion between the Prorocentrum micans diet and ingestion (Table 1) or after 96 h (e.g. Day 5). the Alexandrium strains were due to chemical stimuli produced by the latter, P. micans cells were suspended

Combination diets Table 4. Mean percentage of larvae that show ingestion of indicated 1:1 ratio combinations of prey types for specified When Prorocentrum micans was presented to larvae crab species. See Table 1 for definitions of diet treatments of all 3 species of brachyuran crabs in a 1:1 ratio with any of the 3 Alexandrium strains, nearly 100% of the Crab species Diet Treatment Time interval Pm/At115 Pm/Af1719 Pm/At118 larvae ingested algal cells at all observation times (Table 4). Furthermore, direct observations revealed Cancer oregonensis that in all cases both algal species were being ingested. 390100 100 To further elucidate this observation, larvae of Hemi- 6909590 12 90 90 90 grapsus oregonensis were fed diets consisting of a 24 95 90 90 range of Prorocentrum micans:Af1719 ratios as described 48 100 88 90 earlier. Incidence of ingestion was observed at 8 h. The 96 100 100 90 percentage of larvae feeding began to decline at a Cancer magister 3 100 100 100 ratio of 1:10 (Fig. 1), dropping to less than 50% feeding 6 100 95 90 at a ratio of 1:20, with less than 5% feeding at a ratio of 12 85 90 90 1:30. Again, in all cases where ingestion was noted, 24 100 90 90 both algal species were present in the guts. 48 100 88 90 96 100 100 90 When Hemigrapsus oregonensis larvae were raised Hemigrapsus oregonensis in laboratory culture on the 1:1 ratio of Prorocentrum 390100 90 micans:Af1719, there was a significant delay in mortal- 690100 100 ity compared with an unfed treatment or with larvae 12 90 100 90 fed only Af1719 (Table 5; ANOVA, p < 0.05; Tukey’s 24 95 100 100 48 100 90 100 HSD, p = 0.05). As anticipated from the ingestion ob- 96 100 100 100 servations, there was no delay in mortality on the latter 160 Mar Ecol Prog Ser 222: 155–162, 2001

120 Ingestion of plastic beads

100 When Hemigrapsus oregonensis larvae were given plastic beads alone, very few fed on them (Table 7). 80 However, when the beads were provided in combination with Prorocentrum micans cells at a ratio of 1:1 60 (total particle concentration of 1000 ml–1), 58% of the

% Feeding 40 larvae ingested both particle types. Furthermore, when beads were presented to larvae in combination 20 with Af1719 cells or suspended in either P. micans or Af1719 exudate, there was low incidence of ingestion. 0 As was the case with Alexandrium cells, plastic beads 1:1 1:2 1:5 1:10 1:20 1:30 were ingested most readily when combined with P. Ratio of P. micans : A1719(1000mlÐ1) micans cells. Fig. 1. Percentage of Hemigrapsus oregonensis larvae feeding on indicated ratios of Prorocentrum micans:Alexandrium fundyense 1719 (Af1719) cells. Total cell density is 1000 ml–1 Ingestion of Noctiluca scintillans

Table 5. Mean ± SE day of death for Hemigrapsus oregonen- Noctiluca scintillans cultured on either Prorocen- sis larvae fed the indicated diets. There were significant dif- trum micans or Af1719 were readily ingested by Hem- ferences among treatments (ANOVA, p < 0.05). Shared letters igrapsus oregonensis larvae. When the 2 diet treat- indicate no significant differences (Tukey’s HSD, p = 0.05). See Table 1 for definitions of diet treatments. n: numbers of ments were offered to larvae in combination, 90% of larvae. Total cell counts at 1000 ml–1 except Pm the Af1719-fed and 93% of the P. micans-fed N. scintillans were ingested, with no significant difference Diet Mean day of death n Tukey’s HSD between treatments (Student’s t-test on arcsine transformed data, p > 0.05) Af1719 5.3 ± 0.15 36 a Unfed 5.8 ± 0.25 36 a Pm/Af1719 (1:1) 8.2 ± 0.46 36 b Pm (500 ml–1) 9.2 ± 0.47 34 b DISCUSSION

Virtually all larvae of all 4 species of crabs ingested Table 6. Mean percentage of larvae that showed evidence of Prorocentrum micans, while there was almost univer- ingestion of indicated prey-exudate combinations at each observation time. See Table 1 for definitions of diet treat- sal failure of larvae to ingest any strain of Alexan- ments. Ex: exudates. n = 3 in each case drium. This indicates some level of discrimination in feeding on algal cells by crab larvae. Crab species Diet For larvae to discriminate among prey, they must Time interval At115/ Af1719/ Pm/ Pm/ respond to prey-associated cues, presumably either Pm Ex Pm Ex At115 Ex Af1719 Ex chemical or mechanical stimuli. The mechanism by Cancer oregonensis which crab larvae ingest captured protist cells is not 12 0 0 100 90 well described. Sulkin et al. (1998a) described bra- 24 0 0 100 90 chyuran larvae as ‘encounter feeders’; that is, larvae Cancer magister 12 0 0 100 100 grasp prey particles, both living and non-living (detri- 24 0 0 100 90 tus), as both predator and prey move through the

Table 7. Mean percentage of Hemigrapsus oregonensis lar- in freshly prepared exudate of the non-toxic At115 and vae feeding on plastic beads and algal mixtures after 8 h of the toxic Af1719 as described above. In both Cancer exposure to the specified diets. See Table 1 for definitions of spp., incidence of P. micans ingestion was close to diet treatments. Ex: exudate. n = 3 in each case 100% in the exudates of the 2 Alexandrium strains (Table 6). When Af1719 and At115 cells were sus- Diet treatment Feeding (mean %) SE pended in P. micans exudate, there was no incidence Beads alone 5 0.11 of ingestion (Table 6). Thus, the stimulus to feed on P. Beads + Pm 58 0.23 micans is not affected by the presence of Alexandrium Beads + Af1719 3 0.26 Beads + Pm Ex 12 0.43 exudates, nor are Alexandrium cells rendered more Beads + Af1719Ex 8 0.21 palatable by the presence of P. micans exudate. Hinz et al.: Larval crab ingestion of protistan prey 161

water. Prey capture by crab larvae does not appear to types. While this scenario seems the most likely, the require use of their well-developed eyes (Harvey & results do not rule out the possibility that Alexandrium Epifanio 1997, Sulkin et al. 1998a) and occurs over a cells do produce a rejection cue that is readily over- wide range of prey sizes (from at least 5 to 250 µm; ridden by the presence of a positive cue produced by Lehto et al. 1998). Larger prey typically are grasped by another alga. In either case, however, the positive the endites of the maxillules and pushed toward the stimulus dominates the process. mandibles where they are masticated and ingested Clearly, the failure of larvae to ingest non-toxic as (Crain 1999). It is possible that the protist Noctiluca, well as toxic strains of Alexandrium precludes the pos- being as large as the largest metazoan prey typically sibility that Alexandrium-produced toxin prevented used in laboratory culture of crab larvae, are handled ingestion of cells by larvae. Furthermore, placing lar- in a similar way. Crain (1999) described capture and vae in exudate freshly prepared from either a favor- handling of the smaller Prorocentrum micans cells by able or an unfavorable prey had no effect on their the anomuran crab Placetron wosnessenskii as a ‘fling ingestion behavior. The results thus do not support the and clap’ motion (as described by Koehl & Strickler role of a water-borne chemical stimulus, either positive 1981) in which the mouthparts first are swung rapidly or negative. It is possible, however, that the presumed outward, drawing the cell toward the mouth, then are cue on the prey cell surface is chemical in nature. rapidly drawn in, enclosing the cell. A similar activity Larvae of Cancer oregonensis did ingest Alexan- was noted in the present study when larvae ingested drium fundyense cells, at least initially. Because C. P. micans, but not when they were presented with Ale- oregonensis did not ingest either non-toxic or toxic xandrium spp. In the latter case, larvae pushed the strains of A. tamarense and because other crab species, Alexandrium cell toward the mandibles, which pulled including the congener C. magister, did not ingest A. the cell in toward the mouth, then pushed it away. fundyense, this result appears to be a species-specific Thus, larvae appeared to require intimate contact phenomenon for either the predator, the prey, or both. with the target cell before ingestion was stimulated. Cancer oregonensis has been shown to be more nutri- This is consistent with the lack of stimulus provided by tionally flexible than its congener (Sulkin & McKeen the exudate and with the speculation by Teegarden & 1999) and may be able to consume a wider variety of Cembella (1996) that copepods make prey ingestion prey. It is interesting to note, however, that the per- decisions by recognizing cues either on the surface of centage of larvae that showed evidence of having the cell or in an associated fluid microzone around the ingested this toxic strain declined over time. Presum- cell. Both Alexandrium tamarense and Prorocentrum ably, the presence of toxin or some other factor micans have lectin molecules on the cell surface (Hori resulted in larvae ceasing to ingest this prey after a et al. 1996). It is possible that a lectin or other mole- period of exposure. Unfortunately, there is no non- cule(s) associated with the cell surface play a role in toxic strain of A. fundyense available to rigorously test prey discrimination on the part of larvae. the role played by the toxin in this behavior. The nature of a prey-associated ingestion cue might Given the absence of ingestion of Alexandrium cells be positive (stimulating ingestion) or negative (stimu- by the larvae, it is not surprising that this diet did not lating rejection). The preponderance of evidence sug- support development and that there was no delay in gests that a positive stimulus was controlling ingestion mortality compared with unfed controls in either Can- of protists in the present study. This conclusion is sup- cer species. Inducing Hemigrapsus oregonensis larvae ported by the observation that while Alexandrium spp. to ingest the toxic strain of A. fundyense by combining cells were not ingested, both Prorocentrum micans and it with Prorocentrum micans neither accelerated nor Noctiluca scintillans were, the latter whether it had delayed mortality compared with presenting the lar- previously been fed either P. micans or Af1719. Fur- vae with a sub-optimal dosage of P. micans alone. This thermore, larvae ingested all 3 strains of Alexandrium, suggests that ingestion of toxic strains of an alga that as well as inert plastic beads when they were com- does not itself stimulate predation had no deleterious bined with comparatively few P. micans cells. Thus, effect on the larval predator. Only in the anomuran Alexandrium cells did not appear to produce a cue that facultative planktotroph Rhinolithodes wosnessenskii induced rejection on the part of the larval predator, did extended exposure to toxic Alexandrium have an since they were ingested when combined with P. effect. However, even in this case, results were mixed; micans cells or when inside the food vacuoles of N. a reduction in mortality was combined with a delay in scintillans. The ingestion of both Alexandrium cells development compared with the unfed control. and plastic beads when each was combined with P. Larvae in nature are almost always faced with a micans further suggests that once feeding is induced complex mixture of prey. Feeding is likely to occur if by a positive stimulus, such as that produced by P. the mixture includes sufficient quantities of prey that micans, larvae cease to discriminate between cell produce positive stimuli for ingestion. Moreover, if the 162 Mar Ecol Prog Ser 222: 155–162, 2001

mixture does include sufficient quantities of such prey, Hartman M, Sulkin S (1999) Effects of prior exposure to petro- larvae are also likely to ingest prey particles that do leum hydrocarbon contamination during brooding on the not provide this positive stimulus. This lack of discrim- subsequent larval development of the brachyuran crab Hemigrapsus oregonensis.J Crustac Biol 19:690–698 ination in prey selection when favorable prey are pres- Harvey EA, Epifanio CE (1997). Prey selection by larvae of ent may promote rapid ingestion of available particles the common mud crab Panopeus herbstii Milne-Edwards. when larvae encounter dense patches of prey in an J Exp Mar Biol Ecol 217:79–91 otherwise sparse prey environment. Haynes EB (1984) Early zoeal stages of Placetron wosnessenskii and Rhonolithodes wosnessenskii and review of Litho- It also seems likely that larvae do not feed when they did larvae of the northern North Pacific Ocean. Fish Bull find themselves in dense aggregates of prey domi- 82:315–324 nated by unfavorable and possibly toxic blooms of Hori K, Mimuro M (1996) Lectin-like compounds and lectin algae. However, even harmful algal blooms, although receptors in marine microalgae hemagglutination and consisting of only a few species at the bloom center, do reactivity with purified lectins. J Phycol 32:783–790 Koehl M, Strickler R (1981) Copepod feeding currents: food sustain a more diverse community, including het- capture at low Reynolds number. Limnol Oceanogr 26: erotrophic protists, at the periphery (Turner & Tester 1062–1073 1997). Thus, a prey field may develop at the periphery Lehto J, Sulkin S, Strom S, Johnson D (1998) Protists and of such an algal bloom that is likely to be a favorable detrital particles as prey for the first larval stage of the brachyuran crab, Hemigrapsus oregonensis. J Exp Mar feeding environment for crab larvae. Biol Ecol 230:213–224 Crab larvae, while they are omnivores that ingest a Levine DM, Sulkin S (1984) Nutritional significance of long- wide variety of prey types including metazoans, pro- chain polyunsaturated fatty acids to the zoeal develop- tists, and microbially colonized detrital particles, nev- ment of the brachyuran crab, Eurypanopeus depressus. ertheless appear to discriminate among prey. At least J Exp Mar Biol Ecol 81:211–223 Paul AJ, Paul JM, Shoemaker PA, Feder HM (1979) Prey con- for such small prey as protists, the discriminatory centrations and feeding responses in laboratory-reared behavior seems to require testing of individual cells. It stage-one zoeae of King crab, Snow crab, and Pink has yet to be determined whether there is a relation shrimp. Trans Am Fish Soc 108:440–443 between the presence of a positive prey-associated cue Paul AJ, Paul JM, Coyle K (1989) Energy sources for first-feeding zoeae of king crab Paralithodes camtschatica (Tilesius) and the qualitative value of that prey as a nutritional (decapoda, Lithodidae). J Exp Mar Biol Ecol 130:55–69 source. Staton J, Sulkin S (1991) Nutritional requirements and starvation resistance in larvae of the brachyuran crabs Sesarma cinereum (Bosc) and S. reticulatum (Say). J Exp Mar Biol Acknowledgements. This manuscript includes data submitted Ecol 152:271–284 in partial fulfillment of the Master of Science degree at West- Sulkin S (1975) The significance of diet in the growth and ern Washington University by the first author. The first author development of larvae of the blue crab, Callinectes was supported by graduate assistantships provided by the sapidus, under laboratory conditions. J Exp Mar Biol Ecol Shannon Point Marine Center and by assistance from NSF 20:119–135 Grant No. OCE-9729316. The fourth author was supported by Sulkin S, McKeen G (1999) The significance of feeding his- NSF Grant No. OCE-9731144. tory on the value of heterotrophic microzooplankton as prey for larval crabs. Mar Ecol Prog Ser 186:219–225 LITERATURE CITED Sulkin SD, van Heukelem WF (1980) Ecological and evolu- tionary significance of nutritional flexibility in planktonic Anger K, Dawirs R (1981) Influence of starvation on the larval larvae of the deep sea red crab Geryon guinquedens and development of Hyas araneus (Decapoda; Majidae). Hel- the stone crab Menippe mercenaria. Mar Ecol Prog Ser 2: gol Wiss Meeresunters 34:287–311 91–95 Crain JA (1999) Functional morphology of prey ingestion by Sulkin SD, Blanco A, Chan J, Bryant M (1998a) Effects of Placetron wosnessenskii schalfeew zoeae. Biol Bull 197: limiting access to prey development of first zoeal stage of 207–218 the Brachyuran crabs Cancer magister and Hemigrapsus Damkaer D (1977) Initial zooplankton investigations in Prince oregonensis. Mar Biol 131:515–521 William Sound, Gulf of Alaska and Lower Cook Inlet. In: Sulkin S, Lehto J, Strom S, Hutchinson D (1998b) Nutritional Environmental assessment of the Alaskan continental role of protists in the diet of first stage larvae of the shelf: annual reports of the principal investigators for the Dungeness crab Cancer magister. Mar Ecol Prog Ser 169: year ending March, 1977, Vol. 10. Receptors—Fish, Lit- 237–242 toral, Benthos. RU-425. US Department of Commerce, Teegarden GJ, Cembella AD (1996) Grazing of toxic dinofla- Boulder, CO, p 137–274 gellates, Alexandrium spp., by adult copepods of coastal Gifford DJ (1985) Laboratory culture of marine planktonic Maine: implications for the fate of paralytic shellfish oligotrophs (Ciliophore, Oligotrichida). Mar Ecol Prog Ser toxins in marine food webs. J Exp Mar Biol Ecol 196: 23:257–267 145–176 Hansen PJ (1989) The red tide dinoflagellate Alexandrium Turner JT, Tester PA (1997) Toxic marine phytoplankton, zoo- tamarense: effects on behavior and growth of tintinnid plankton grazers, and pelagic food webs. Limnol ciliate. Mar Ecol Prog Ser 53:105–116 Oceanogr 42:1203–1214

Editorial responsibility: Michael Landry (Contributing Submitted: October 30, 2000; Accepted: March 22, 2001 Editor), Hawaii, Honolulu, USA Proofs received from author(s): October 26. 2001